What Semaglutide Costs Around the World, and Why the Gap Is So Large

I spent a good part of this week converting the price of one molecule into five currencies, which is not a hobby anyone chooses. If you came here asking about semaglutide cost, the plain answer is that there is no single number. There is a price in India, a different price in Canada, a much later generic date in the US, and the gaps between them come from patents, extensions and who is doing the negotiating. They do not come from what the stuff costs to make.

Some numbers. In India, before generics, Ozempic cost about ₹8,800 to ₹11,175 a month (roughly $100 to $130) and Wegovy up to ₹16,400 (roughly $190). After more than 40 generic brands launched on March 21, 2026, generics sold for ₹1,800 to ₹4,200 a month (roughly $20 to $50). In Canada, Sandoz projected generics would take monthly costs from about CAD 1,000 to CAD 300 or less (roughly $720 down to roughly $215). In the US, litigation settlements point to generic entry around 2032, and my facts do not include a settled US brand price to set beside these, so I will not make one up. A 2024 JAMA report put the cost of producing a month’s supply of semaglutide powder at 7 cents.

Why the price has nothing to do with the cost of making it

Start with the reasonable version. A company spends years inventing a drug. Once invented, the drug is cheap to make: about 7 cents of powder a month, and Chinese ingredient makers could reportedly sell it to compounders for up to about 49 cents and still collect seven times their cost. (The pen, the refrigeration, the regulators, etc., cost more than the powder, in fairness to everyone.) So the price of the drug is not the cost of the drug. The price pays for the invention.

That is the patent bargain, and it’s fine. The government lets you be the only seller for 20 years, you charge a lot, and then the monopoly ends and the price drifts down toward the 7 cents. Novo Nordisk filed the core semaglutide patent on March 20, 2006, which gives a baseline expiry of March 20, 2026.

Commercially, the bargain worked. IQVIA calls semaglutide (Ozempic and Wegovy combined) “the world’s second-best-selling prescription medicine,” with $26 billion in sales in 2024, growing about 40% a year. That is a lot of inventing paid for.

How each country gets its own patent clock

Step two. Patents are national. So instead of one monopoly ending on one day, you get a monopoly that ends in different places on different days, and the price of a single molecule becomes a function of geography.

The rough schedule: an early wave in 2026 (India, China, Brazil, Canada, South Africa), a mid wave around 2031 (the UK, Germany, France, Italy, Spain and Japan), and a late wave in 2032 to 2033 for the US. The difference is extensions. The European Union granted a Supplementary Protection Certificate adding five years, to 2031. The UK one runs from March 19, 2026 to March 19, 2031.

In the US, the compound patent got what the law calls “patent term restoration,” a 1,040-day extension to December 5, 2031. Translated: the government agreed that time spent waiting for approval did not really count, and handed back about two years and ten months. Novo Nordisk’s own 2025 Form 20-F gives the US compound expiry as “2032,” with secondary patents running to 2038 and one Wegovy patent to May 13, 2040.

The early-wave countries are not uniformly early, either. Brazil’s regulator, ANVISA, refused Novo’s extension request and approved the first generic in May 2026 for diabetes pens, though one pricing guide projects generic pens reaching Brazilian pharmacies around 2028 to 2030. China’s compound patent expired, but data protection runs to about April 2027, with at least 17 Chinese manufacturers in Phase III or pre-market stages. In India, a tablet formulation patent runs to 2031, so the pill version stays protected even though the injection does not.

So a patient in Mumbai and a patient in Ohio take the same molecule on very different timetables. Same chemistry. Different calendars.

What semaglutide cost does when 40 brands show up

India is the clean experiment. The patent lapsed on March 20, 2026. On March 21, more than 40 brands launched. Within 10 days, Novo cut Ozempic by 36% and Wegovy by 48%. By April, generics had taken 50% of India’s semaglutide sales by value.

Note what did not change in those 10 days: the cost of making the drug. The powder still cost what it cost. What changed was the number of sellers. That was the whole difference.

It must have been an exhausting week and a half for Novo’s pricing team, discovering under such pressure that the drug could be nearly half as expensive. Very fast work. Impressive, really.

How a country loses its patent over a fee

Step three is where it gets weird. Patents also have to be maintained, which means paying renewal fees, which means somebody has to remember. In 2019, Novo did not pay a CAD 250 (roughly $180) maintenance fee on Canadian patent CA 2601784, and the patent lapsed permanently. That also voided a supplementary protection certificate that would have run to March 20, 2028. Health Canada approved Dr. Reddy’s generic on April 28, 2026, the first G7 country to do so, in what is reportedly the world’s second-largest semaglutide market.

In a heist movie the vault has lasers and a guard. Here, the vault was protected by an invoice. (I write about patent maintenance fees for a living, which is the sort of sentence that ends conversations at dinner.)

The patent firm Reddie & Grose titled its write-up on this “Deliberate or a Mistake?”, and I don’t know. The obvious reading is that somebody missed a payment, the way people miss payments, possibly while someone in the office was saying, “hey, did the Canada one go out?” (a line I have invented). The other reading is that a company with dozens of patents on one drug decided this particular one was not worth $180 in 2019, before anyone knew how big the drug would become. The facts here do not say which. Either way, a fee smaller than a monthly prescription set the generic date for a whole country.

Where reference pricing takes the price next

Now the last step, and the dumb endpoint. Suppose countries stop setting prices on their own and start pointing at each other. Here is an invented scenario. Country A says it will pay the lowest price charged anywhere in a basket of rich countries. Country B, which is in the basket, says the same thing. So does C. Now every price is defined as some other country’s price. The company, for instance, could respond by raising prices in B so A’s reference goes up, or by delaying launch in B so there is no B price at all. Push it far enough and the drug has no price, only a circle of countries each waiting to see what the others pay. Which is sort of where the arguments already go.

The real world is partway down that road. Ozempic, Rybelsus and Wegovy were picked for the second round of Medicare drug price negotiation under the Inflation Reduction Act. One study re-estimating Medicare spending on semaglutide for 2026 to 2035 found base-case savings of $463 million under the negotiated price, rising to $2.63 billion under “Most Favored Nation” pricing combined with generic entry assumptions. Most Favored Nation is the circle: tie what Americans pay to what other countries pay.

The point is that foreign prices stop being foreign. Under a regime like that, an Indian price cut or a Canadian generic is no longer just a local event; it is potentially an input into a US number. Exactly which countries count, and how the formula works for semaglutide, is not settled in what I have here.

If the basket ends up including Canada, then some piece of what Medicare pays for semaglutide will trace back to a roughly $180 fee nobody paid in 2019.